ZipCrawl County-Level Data File

Version: v1.0
Last Updated: November 23, 2025
Geography: U.S. Counties (FIPS)
Source Datasets: ACS 5-Year Estimates, ZipCrawl curation pipeline

This dataset is available as a standalone file, or as part of our curated U.S. Geography bundle.

• County Curated Dataset (U.S.)
• U.S. Geography Curated Bundle

You can also view data for a specific county here: County Lookup (U.S.)

Overview

This dataset provides curated demographic, economic, housing, and geographic indicators for all counties and county-equivalents in the United States. This is based on U.S. Census county equivalents: County FIPS. FIPS is a numeric identifier for each County or equivalent (such as a Parish, in Louisiana).

ZipCrawl uses raw ACS estimates with and applies essential data-quality safeguards to ensure that values are interpretable, consistent, and appropriate for analytical use.

How to Use This Dataset

Understanding Null Values

null in this dataset never means zero.

A null indicates:

• unreliable ACS estimate (high margin of error), or...
• data that was not available in the source data set.

Do not treat null as zero in analysis unless explicitly intended.

Join Logic

Counties are stable and uniquely identified by: county_fips (string)

Recommended Use Cases

• ZIP-level segmentation
• Data enrichment for consumer, real estate, or policy analytics
• Modeling markets, demographics, or economic conditions
• Geographic clustering

Common Pitfalls (Read This!)

⚠️ Pitfall 1 — Misinterpreting NULL values

NULL = “statistically unreliable,” not “zero.”

⚠️ Pitfall 2 — Percentages may not sum to 100%

ACS universes differ across concepts (age, employment, households, etc.).

Methodology Summary

ZipCrawl applies a small number of targeted adjustments to ensure dataset quality while preserving ACS integrity.

Margin-of-Error (MOE) Filtering

A value is set to NULL when: MOE ≥ Estimate * 0.5 AND the target value ≥ 100

This follows best practices to ensure that unreliable estimates on small population sizes are not included in the final data. On sample sizes closer to zero, even with a large MOE, we can be more confident the true value is close to zero. We never suppress these columns to null where a value exists, due to their outsized importance:

• total_population We use a higher threshold (1000) on ethnic breakdowns, since they collectively add up to one. We only want to suppress these if they are grossly unreliable.

Universe Alignment for Percentages

ACS percent calculations use different universes (e.g., population 3+, population 18+, civilian labor force, household population).
ZipCrawl ensures:

• numerator and denominator universes match
• no percentage exceeds logical bounds
• questionable ratios are suppressed or corrected

Range Validation

Percent fields are validated to ensure: 0 ≤ pct ≤ 1

Values outside this range indicate universe mismatch or erroneous sources and are suppressed.

Field Groups (High-Level Structure)

Fields are organized into the following groups:

Geography & Identification

• county_fips
• state

Population

• total_population
• age structure
• sex ratios

Race & Ethnicity

• Standard ACS race and Hispanic origin distributions.

Education

• educational attainment
• enrollment
• universe-aligned percentages

Economics

• income
• poverty
• employment & labor force metrics

Housing

• occupancy
• tenure
• rent & costs
• household characteristics

Versioning and Stability

This dataset is updated annually following the Census Bureau's ACS 5-Year release schedule. There may be periodic updates as other sources are added or updates.

Each release includes:

• updated ACS estimates
• stable ZipCrawl methodology
• version notes summarizing methodological changes (if any)

Changes to methodology are expected to be rare and will be documented clearly.

Contact

For questions or feedback:

ZipCrawl Support Team
support@zipcrawl.com